Cheap plastic Compass
 

Posted by Kris Verdeyen at 03/28/2001 5:26 PM EST


Engineer on team #118, Robonauts, from CCISD and NASA - Johnson Space Center and Friends.


In Reply to: Erm....
Posted by Kevin Sevcik on 03/28/2001 3:14 PM EST:



I've not actually done any work with the kit gyro, but some of these solid state sensors now can do really amazing things. It should be able to sense bumps, though, because it's reading actual yaw, as opposed to using dead reckoning (where it would be screwed up if the wheels were to slip). The trouble would probably come into play when we went over a bridge while turning. It would integrate the yaw change wrt the robot, and wouldn't have any way of translating that to an actual change on the field, I think. Actually, it might all work out in the end.

But that compass thing was interesting. - Why not hook up a cheap plastic compass (from SPI, of course) to a pot, and there we have our sensor. :)

Re: Has anyone succesfully programed the Gyro?
 

Posted by Edward Gilchrest at 03/30/2001 4:04 PM EST


Engineer on team #38, Nonnebots, from Nonnewaug High School and 9th Wave, Inc..


In Reply to: Has anyone succesfully programed the Gyro?
Posted by Anthony S. on 03/27/2001 8:50 PM EST:



: I'm a student from team #442, we initially planned to use the gyro to balance but on night before shipping decided it wasn't going to work therefore we built our famous "outrigger" A.K.A "pogo-stick". So I was just wondering if anyone knew how to work it, we tried forever to program it but never go it to. Please let me know if you did, I'm very anxious to see if anyone was able to get it working.

I am the president of 9th Wave Design, the sponsor for the Nonnebots, team #38. We have designed a gyro program that works successfully and consistently. The program is fairly simple. When we started designing the software we analyzed all the masses, wrote a program based on a PID algorithm (proportional, integral, derivitive) but found the loop speed to be too slow to properly manage it.
We then wrote a program that looks for the ramp to start tipping, immediately moves the robot backwards while calculating the angle and stops the pwm at a prescribed angle. This program also has a compensation for battery voltage, since the speed at which a particular pwm drives the robot is heavily dependent on battery voltage.
A very important aspect of the program is the fact that the operators clear the gyro counters and set a zero point while first moving up the ramp so errors don't add up.
The program is designed to balance the ramp without overshoot so it does not "close the loop" so to speak as a true PID algorithm would.
This program has worked well with one goal, two goals, and goals filled with balls without having to make any changes.
If anyone would like more info feel free to email me at egilchrest@9thwave.com